[ratelimitsteve]

I don't think any of this review is unfair. Shipping a broken product and then covering the fix with your warranty is better than shipping a broken product and telling the consumer to get bent, but worse than shipping a working product. A product that advertises fewer features and delivers them is better than a product that advertises more features and doesn't deliver them. Even in the case of the CO2 sensor, it may be your opinion that a CO2 sensor is critical and can't be done without, but the review is for the feature set at the price point. A device that does fewer things can be a better device depending on price point and, again, reliability.

If, however, I concede the author's idea that reviews must have objective criteria, methodology and standards in order to be taken seriously then I'd like to propose the first objective criterion: broken out-of-the-box === not recommended.

edit: evidently the device failed after a few months. This doesn't change my final opinion, which is in total agreement with the review, but it deserves to be mentioned because I was incorrect in my facts. For my fellow JS devs, I'm standing by broken out-of-the-box === not recommended, and adding broken within a few months of installing == not recommended.

[beAbU]

I did not pick up that the device was broken when it was sent out from the article. It failed "After using the monitor for a few months".

[thmsths]

That is still not good, devices tend to follow the bathtub curve, so if it took only a few months for it to fail, they either got really unlucky OR the device is highly unreliable.

[jesol]

I don't have hard data, but I own 3 air gradients and they've been going strong for years. I don't think it's unreasonable to think they just got unlucky

[ratelimitsteve]

>they just got unlucky

then let me ask you: as an ethical person trying to write a review, how do you handle that situation? It seems like that's an angle to this that we're not exploring, and that's whether the review is epistemologically justified rather than whether it's objectively correct. The way I see it, as a reviewer I get a product that fails well sooner than I expected it to and I have three choices:

1) Don't report the failure

2) Report the failure

3) Report the failure but try to contextualize it (basically, trying to solve for sure whether they got unlucky or not)

1 is obviously unethical, and 3 seems like it's well outside the scope of a reviewer's duty (and could be seen as carrying water for a particular brand. after all, do you think the person who wrote the OP would be okay with it if his product's failure was considered typical but another product's failure was determined to be atypical, regardless of the truth?). The only ethical approach is to report what happened, and not speculate as to cause.

[thmsths]

Your 2nd option is what I would like to see. And I do agree that it is not perfect, but as you pointed out, it is the least bad option.

[BizarroLand]

Do what Linus Tech Tips has done (when they're doing it right): report that they got the device, tested it for a while, and it broke, so what they currently know about it is not the full picture, and that they might revisit the product when they have confirmed that they've worked out the kinks with the device in question.

Easy

[ratelimitsteve]

That feels like exactly what the article did. They disclosed that the thing broke and on what time frame, alongside details about the degradation of functionality. As far as say that this isn't the full picture, is it definitely not? If we owe someone the benefit of the doubt if their device breaks while being reviewed do we owe the reader the deficit of the doubt if it doesn't break? What if they just got lucky?

[mrgoldenbrown]

if your sample size is 1 of each product, your review is likely going to be unfair to someone. Especially as you test more and more models.

[burnt-resistor]

This. And what if the reviewer or someone else smashed the box unreasonably?